Projective multiresolution analyses for $L^2(R^2)$

Judith A. Packer,Marc A. Rieffel
DOI: https://doi.org/10.48550/arXiv.math/0308132
2004-03-10
Abstract:We define the notion of "projective" multiresolution analyses, for which, by definition, the initial space corresponds to a finitely generated projective module over the algebra $C(\btn)$ of continuous complex-valued functions on an $n$-torus. The case of ordinary multi-wavelets is that in which the projective module is actually free. We discuss the properties of projective multiresolution analyses, including the frames which they provide for $L^2(\brn)$. Then we show how to construct examples for the case of any diagonal $2 \times 2$ dilation matrix with integer entries, with initial module specified to be any fixed finitely generated projective $C(\mathbb T^2)$-module. We compute the isomorphism classes of the corresponding wavelet modules.
Functional Analysis,Operator Algebras
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to introduce the concept of "projective" multi - resolution analysis in the multi - resolution analysis (MRA) framework in order to extend the traditional wavelet theory. Specifically, the author explores how to generalize the classical multi - resolution analysis to a more general situation, that is, the initial space corresponds to a finitely generated projective module on \(C(T^n)\), rather than the usual free module. This generalization allows the use of a wider range of mathematical structures to construct wavelets and the corresponding normalized tight frames, thereby providing new construction methods for \(L^2(\mathbb{R}^n)\). ### Specific problems include: 1. **Define projective multi - resolution analysis**: - Traditional multi - resolution analysis requires that the initial space \(V_0\) be the linear span space generated by the integer translations of one or more scaling functions. In this paper, the author relaxes this condition and only requires that \(V_0\) be the space of the inverse Fourier transform of a finitely generated projective module on \(C(T^n)\). 2. **Construct non - trivial projective multi - resolution analysis**: - The author shows how to construct non - trivial projective multi - resolution analysis for any diagonal \(2\times2\) dilation matrix and the space of continuous sections of any complex vector bundle. This solves the problem of whether there exist non - trivial projective multi - resolution analyses. 3. **Determine the isomorphism classes of wave modules**: - The author calculates the isomorphism classes of the wave modules corresponding to the constructed projective multi - resolution analyses. These wave modules are free modules in some cases and not in others. 4. **Apply to high - dimensional cases**: - Although the main discussion is about the two - dimensional case, the author points out that their construction can be extended to higher dimensions, and for some specific dilation matrices (such as diagonal matrices), projective multi - resolution analyses have been successfully constructed. ### Mathematical formula representation: - **Projective module**: \( V_0 \) is a finitely generated projective module on \( C(T^n) \). - **Dilation matrix**: Let \( A \) be an \( n\times n \) invertible matrix, and the modulus of all its eigenvalues is strictly greater than 1. - **Unitary operator**: \( D \) is a unitary dilation operator on \( L^2(\mathbb{R}^n) \), defined as \( (D\xi)(t)=\delta^{-1}\xi(At) \), where \( \delta = |\det(A)|^{-1/2} \). - **Inner product**: In the frequency domain, the \( C(T^n) \)-valued inner product is defined as \( \langle \xi,\eta\rangle_A(x)=\sum_{p\in\mathbb{Z}^n}(\overline{\xi}\eta)(x - p) \). Through the solutions to these problems, the author shows how to generalize the classical wavelet theory in a broader algebraic framework, thereby providing new tools and methods for signal processing, image compression and other fields.